Identification of urban masses using night Remote Sensing (Case study: Ahvaz)

Rapid, irregular, and scattered urban development is the source of many urban problems. Understanding these issues demands a deep understanding of the distribution of the urban spatial structure. The urban mass is the occupied part of the city, whose accurate knowledge, helps the optimal urban management and can be a guidance light for the use of urban spaces in the future. For this purpose and with the aim of accurate and quick identification of the Ahvaz urban mass as the study area, the images taken by the NPP-VIIRS sensor at night are assumed as the base of this research. Moreover, the other factors such as improved vegetation, auxiliary points containing field information, earth surface temperature and geometric mean of the image, are used to prepare three indices HSI, HSI-POI (HP) and HSI-POI-LST (HPL) used in ENVI and ArcGIS software. Then, a support vector machine was used to identify the type of land in the built area using 1000 points. Next, the threshold limit corresponding to the highest overall accuracy and the kappa coefficient of each index were calculated. The general results showed that compared to other indices, HPL shows natural and social characteristics, and the output of this method provides higher spatial accuracy, more spatial details of the built-up area, and also more choices for evaluating small-scale urban areas. Also, The highest and lowest net area of the urban mass belongs to regions 3 and 1, which includes more than 56 and 34 percent of the area of each of these regions.